Transcript from public YouTube captions. Lightly corrected for names; automatic transcription may contain errors.
Source: https://www.youtube.com/watch?v=mMjck0bXMps

[0:01] All right, welcome back to a chat with
[0:03] chat. Today we're joined by Shaurya Jain,
[0:07] founder of Honest HVAC Marketing Agency.
[0:11] Uh he got started in digital marketing
[0:13] back in 2013. And today we're going to
[0:16] be exploring how HVAC businesses work
[0:19] when customers ask AI for
[0:21] recommendations, what Shaurya is learning
[0:25] from his work, and how contractors are
[0:28] adjusting to the visibility with AI
[0:31] search and generative engine
[0:32] optimization. Shaurya, welcome to the
[0:35] podcast. Tell us a little bit about your
[0:38] background and what you've been up to.
[0:40] >> Yes. So, first of all, thanks Aaron for
[0:42] inviting me over as being the first
[0:45] guest. So, what I've been up to, I've
[0:47] been up to the same thing as you
[0:49] basically helping HVAC companies and I
[0:53] don't have the advantage of running an
[0:56] HVAC company like you have. But I do
[0:59] have extensive experience in being a
[1:02] marketer. So, I've been a marketer for
[1:04] the last 13 years. And we have been
[1:07] helping HVAC companies with their
[1:09] Facebook ads and local SEO. We have been
[1:13] doing tons of research to see how HA
[1:18] companies are do like doing in the
[1:21] search results in AI how they have set
[1:25] up their websites. I just I yesterday
[1:28] itself I published a new study where we
[1:31] studied like 1,300 HVAC websites to see
[1:34] how ready were they for generating leads
[1:37] like how good were they good were the
[1:41] websites in terms of generating leads.
[1:44] So I'm excited to to talk about those
[1:46] things with you.
[1:49] >> That's fantastic. Now, yeah, I'm very
[1:52] interested in the research that you did.
[1:54] 1,300 websites, that's a pretty good
[1:56] sized sample. How did you go about
[2:00] analyzing that many businesses or that
[2:03] many business websites?
[2:04] >> Yeah. So, I first of all scraped a lot
[2:08] of 1500 or 2,000 HVAC businesses across
[2:11] different cities in the US. And then we
[2:14] created a python script where we
[2:18] asked where we tried to look at
[2:22] different things. So the five different
[2:24] things that we try to look at were let
[2:27] me read those to you. So how fast can a
[2:30] visitor contact the business? How many
[2:34] ways they have to contact the visitor?
[2:36] whether the page earns the trust of the
[2:39] visitor, whether the sites whether the
[2:41] site measures results and whether it
[2:45] works on a phone well enough. So within
[2:48] each of these five categories, there
[2:49] were different things that we looked at.
[2:51] How many ways did listed on the website
[2:55] to be for contacting? Was there a
[2:58] contact form as well? Was there a button
[3:01] to chat with the business? We looked at
[3:05] if the star rating was there, how many
[3:08] reviews the business had? Did they
[3:11] present it on the website? Did they
[3:13] present the actual reviews they have
[3:15] got? We looked at things like sideway
[3:20] whether the font was legible or not on a
[3:24] mobile whether they were using their
[3:27] real photos of their crew or like stock
[3:30] photos because stock photos usually
[3:33] don't really insp like they are not
[3:37] really trustworthy like they don't seem
[3:40] very trustworthy brands that use stock
[3:43] photos. So we looked at a bunch of such
[3:47] stuff and if you want me to like share
[3:50] with you the major findings I I could do
[3:53] that for you.
[3:54] >> Oh yeah. So it sounds like you had a
[3:57] number of different parameters that you
[4:00] were checking these websites against. So
[4:01] do you know how many parameters were
[4:03] there that you were testing?
[4:06] I don't have it at the back of my hand,
[4:10] but you could say if there were like
[4:13] five categories, there would be like
[4:15] five each, five subcategories within
[4:18] each of 25 to 30 parameters.
[4:21] >> Yeah.
[4:21] >> Yeah. Excellent. Okay. And then, of
[4:24] course, I don't want to forget my
[4:26] co-host Chat who's been kind of
[4:28] patiently listening to us over here.
[4:31] Chat, did you have anything to chime in
[4:33] on the conversation so far?
[5:05] All right. So that's a good question. So
[5:08] basically I used your friend Claude
[5:11] cowork to visit each of So first of all
[5:15] we took a screenshot of the homepage of
[5:19] each of the website and then Claude
[5:22] analyze if it was a stock photo or if it
[5:26] was a like a real crew, real trucks, a
[5:31] real company. So that's how we were able
[5:33] to analyze if they were using stock
[5:36] photos or real photos.
[5:40] >> Yeah, that's good. So you're using
[5:42] Claude to analyze the photos and I'm
[5:45] assuming was Claude analyzing the other
[5:48] parameters that you had. You had
[5:49] mentioned that you were running a a
[5:51] Python script. How much of the work was
[5:54] Claude doing versus what Python was
[5:56] looking for?
[5:59] Python the Python script went out and
[6:03] scraped all the data but the data
[6:05] crunching that was done by Claude. So it
[6:08] was able to like I was able to do this
[6:12] entire research in over two days you
[6:15] could say the the only time consuming
[6:18] part was the running the Python script
[6:22] it going out and collecting all that
[6:25] information.
[6:26] >> Yeah. Yeah. Okay. So 1,300 websites
[6:30] checked in 2 days or was it actually
[6:33] more than 1300 or cuz you had mentioned
[6:36] 2,000 I think at the beginning that that
[6:38] didn't work for this test. How did that
[6:41] go?
[6:42] >> Yeah. So basically they were more but
[6:45] some websites don't really allow like
[6:48] they have a cloudflare on installed on
[6:51] them. So the they stop the bots
[6:54] >> from visiting the website. So we had to
[6:57] discard those websites.
[6:59] >> Okay. Okay, that makes sense. And yes,
[7:02] let's get to the meat of it. What were
[7:04] some of your findings from those 1300
[7:06] websites you found you checked?
[7:09] >> All right. So the
[7:11] one of the major findings was that for
[7:14] about 40% of the websites don't have any
[7:18] other way they don't give any other way
[7:21] to the homeowner to contact them other
[7:23] than calling them. So they have no form,
[7:26] they have no booking link, they have no
[7:28] chat, nothing like that. Then only about
[7:32] 28% of the sites re showed their star
[7:36] rating and their actual reviews on the
[7:39] website. So like about 70% of the
[7:42] websites don't even show their reviews
[7:45] which really help with social proof and
[7:51] about
[7:53] only about 20% of the websites had real
[7:57] photos of their crew trucks and the work
[8:00] that they are doing. About 80% didn't
[8:03] have that. So that's also really bad in
[8:06] my opinion.
[8:07] >> Yeah. And and we're not talking about
[8:09] highlevel stuff here. This these are
[8:11] some of the basic, you know, blocking
[8:14] and tackling of building a home service
[8:18] website. Yeah. Not having your reviews
[8:20] on your website and then not having a
[8:23] picture of of your crew. Um were you
[8:27] surprised at the results?
[8:30] Yes, especially for the
[8:33] stock photos thing because if 80% of the
[8:36] businesses HVAC businesses are using
[8:38] stock photos and for me like showing
[8:42] your crew, showing your work that that
[8:47] is one of the biggest lever what one
[8:52] could pull for like getting the trust of
[8:55] the homeowner by showing your real team.
[8:58] So yeah, that was really surprising.
[9:02] >> Yeah. Yeah. And so how do you think that
[9:06] some of those results are impacting
[9:10] those companies and their business? What
[9:13] when you don't do those things, what
[9:15] happens?
[9:18] I think
[9:20] it's killing their conversions. So even
[9:23] if you get someone to visit your
[9:25] website, even if you rank high on
[9:28] Google, someone visits your website, but
[9:31] they don't trust you enough because you
[9:33] are not showing your reviews there, you
[9:35] are not showing the work, the before and
[9:38] after pictures, the testimonials, the
[9:41] video testimonials, your crew, the
[9:44] people that work for you. then you will
[9:47] lose out to people that are doing that
[9:51] even if they are ranking
[9:53] a bit lower on Google because people
[9:57] want to feel safe when they
[10:01] decide to work with an HVAC company.
[10:04] They don't want to work with someone
[10:06] that looks shady.
[10:08] >> Yeah. Yeah. Trust trust is one of the
[10:11] most important factors. You're inviting
[10:14] someone into your home. Yeah.
[10:15] >> Right. And yeah, it's amazing the
[10:19] quality of websites we're still seeing
[10:21] in 2026 now, right? As there's no excuse
[10:25] anymore.
[10:27] >> You don't have to learn how to build a
[10:29] website with WordPress or whatever other
[10:32] complicated builder that that's out
[10:34] there. Um, yeah, there's really no
[10:37] excuse anymore. So, that that's
[10:39] interesting. What inspired you to do a
[10:42] study like this?
[10:45] content. I just wanted to produce some
[10:47] content to produce some thought
[10:49] leadership content. So I basically asked
[10:52] Claude itself that hey give me some
[10:54] ideas about some studies that I could do
[10:56] that wouldn't
[10:58] require too much technical skills and
[11:00] this is one of the studies that I could
[11:03] do without requiring too much scraping
[11:06] too much technical knowledge.
[11:09] >> Excellent. Okay Chat, did you have
[11:11] anything to add to the conversation?
[11:14] Let me check.
[11:15] >> I'm really scared about this.
[11:18] >> ChatGPT chiming in.
[11:19] >> List of things contractors can check.
[11:22] Shaurya, have you tested those changes with
[11:25] a client and measured whether more
[11:27] visitors actually contacted them or
[11:29] booked a job?
[11:32] >> Yeah. So, I haven't got the chance to
[11:35] test it, AB test it because I just
[11:39] published it yesterday. But like
[11:40] intuitively speaking, we all know that
[11:43] if you put your team's photos on the
[11:48] website, if you put the work that you
[11:50] have been doing on your website, your
[11:54] social proof, wherever it is on Yelp, on
[11:57] Google review, if you put it on your
[11:59] website, that will help you with the
[12:01] conversion rate. That is like any
[12:03] marketer can intuitively say that.
[12:09] Yeah, I have to agree. The the actual
[12:12] results would be interesting to see, of
[12:14] course, but there's going to be so many
[12:16] factors because if you're coming across
[12:18] a website that doesn't have pictures of
[12:20] its employees, that doesn't have
[12:23] customer reviews, you're not going to
[12:25] tell go to them and say, "Hey, I know
[12:27] that you're trying to pay bills,
[12:30] and I know, yeah, maybe business is
[12:33] important to you, but I got this test I
[12:35] really want to run, right? Yeah,
[12:37] >> I'm going to change one thing this month
[12:39] and I'm going to change one more thing
[12:41] next month and they're going to go,
[12:42] "What the hell am I paying you for?" You
[12:44] know why?
[12:45] >> Why?
[12:46] >> This is like the bare minimum I would
[12:48] say that we don't even need to think
[12:50] about doing this stuff. It's like common
[12:53] sense now.
[12:54] >> Yeah. Yeah. It's going to happen. It's
[12:56] going to happen. So
[12:59] that research that you were doing, how
[13:02] do you think it applies to uh AI search
[13:06] and generative engine optimization? Have
[13:08] you done a lot of research on that at
[13:11] this point in time?
[13:13] >> Yeah. So
[13:15] with AI, I'm not too sure. So I do know
[13:19] that when AI searches for
[13:23] so when someone searches for HVAC
[13:25] companies in the fan out queries people
[13:29] the the LLMs do look for like
[13:31] testimonials
[13:33] reviews that kind of stuff. So if you
[13:36] put it on your website as well that
[13:39] could help you but I do know that these
[13:42] things help you rank higher in the SERs.
[13:46] So like there's this local SEO expert
[13:49] Joy Hawkins like she is very famous in
[13:53] the local SEO space. So she did
[13:56] presented at a conference where she said
[13:59] that like she had done split testing
[14:02] with her clients where she put real
[14:05] photos of their crew of their team
[14:08] members on their websites. real photos
[14:11] of so let's say if it's a page about
[14:16] HVAC in Los Angeles so and if it's a
[14:22] page let's say targeting a specific
[14:25] neighborhood in Los Angeles so on each
[14:29] of these neighborhood pages she put a
[14:33] specific like picture of the work that
[14:36] the that the crew did in that
[14:39] neighborhood. So, so on all the pages
[14:41] like it the the images were not
[14:44] repeated. They were all they were all
[14:47] what do you say unique
[14:49] >> unique
[14:49] >> and that she said that helped with the
[14:52] rankings that single thing. So
[14:55] >> yeah yeah that's that is interesting as
[15:00] far as your customers go. Do you have
[15:03] customers that are asking about AI
[15:06] search at this point in time? Have they
[15:07] come to you and said, "Hey, how do we
[15:09] start to show up?"
[15:12] >> Yeah, some of them do have asked me and
[15:15] we are now offering this as well. But
[15:19] like it's so difficult to track this
[15:25] thing and also I find it a bit difficult
[15:28] to wrap my head around because it's such
[15:30] a new thing that what to track exactly
[15:34] and how to track it. So we are still
[15:36] learning that thing but yeah clients
[15:39] have been asking about this.
[15:41] >> Yeah it's from what we've been seeing
[15:45] something that I've noticed even on my
[15:47] website is that there's areas where I
[15:51] will not show up in Google for ranking
[15:55] even though I've spent a ton of time
[15:57] dedicating to SEO for certain keywords.
[16:01] But then we'll start getting traffic and
[16:03] inquiries from ChatGPT and we're
[16:07] getting recommended. And it's crazy to
[16:10] me that you could have something that
[16:12] like this that is so new where you're
[16:16] showing up in one area AI search now and
[16:20] you are really struggling to show up in
[16:22] in Google search.
[16:24] >> Yeah, that's true. That's true. I've
[16:25] noticed that like one of my websites
[16:29] were was ranking in ChatGPT and it was
[16:31] sending me tons of leads and it was just
[16:34] a onepage website ranking for just one
[16:37] particular keyword and it had zero DR
[16:41] zero anything but somehow ChatGPT
[16:44] picked it up and started recommending it
[16:46] and it's nowhere to be seen on Google.
[16:49] So
[16:49] >> yeah.
[16:50] >> Yeah, it's really but I think at the end
[16:53] of the day, Google is going to come out
[16:55] as the winner. So we have to we should
[17:00] optimize our website according to Google
[17:04] AI instead of ChatGPT.
[17:06] >> Yeah, a lot of people are predicting
[17:08] that Google's going to eventually end up
[17:10] on top. They have all the data. They've
[17:12] got a huge head start with the search
[17:14] that they've been doing. They seem to
[17:16] have been hesitant to release maybe some
[17:19] of their higher quality or newer models,
[17:21] but they've so much of the traffic right
[17:23] now. I mean, they would really have to
[17:26] flop in to not come out on top.
[17:29] >> Um, and so, yeah, it's been a pretty
[17:32] wild and interesting experience. Now,
[17:36] personal agents seem to be the new rave
[17:40] and what's coming out now. Have you
[17:42] tested any personal agents yourself like
[17:45] Muse or you've got Grockbot or Dots?
[17:49] Anything like that?
[17:50] >> No, I haven't got the chance to do that
[17:53] yet.
[17:54] >> Okay. Yeah, that that seems to be like
[17:57] the next wave. Right now, we're stuck in
[18:01] in AI and agents are getting um are
[18:05] coming out pretty quickly. So, it will
[18:07] be interesting to see where that goes.
[18:09] So,
[18:10] >> yeah.
[18:11] Chat, did you have any questions that
[18:13] you wanted to add in?
[18:16] >> Let me check that quickly.
[18:22] >> That's interesting. Sure am. That
[18:23] onepage website is a great example. How
[18:26] did you establish that those leads came
[18:28] from ChatGPT? Did people tell you or
[18:30] could you actually track their visits
[18:32] through to an inquiry?
[18:35] >> Yeah. So yeah when I got on calls with
[18:39] these people so I asked them like hey
[18:43] where did you find me and some of them
[18:46] didn't remember at all and some of them
[18:48] did and then when I went into ChatGPT
[18:53] so one of the person actually told me
[18:56] the exact thing that they searched for
[18:57] so when I searched for that even on my
[19:00] end then I came to know so at for this
[19:03] particular thing I'm ranking in ChatGPT
[19:07] because no one else was like targeting
[19:09] that keyword and I somehow was. So
[19:13] that's how it happened.
[19:15] And was that an intentional was that the
[19:18] point of that page or was it something
[19:21] else that you were working with and it
[19:24] was just an accidental discovery?
[19:27] >> Yeah, it wasn't intentional at all. I
[19:30] didn't plan to have that website show up
[19:34] anywhere in the SERs or in any chat LLMs
[19:38] because I didn't do any SEO work on that
[19:40] website. It was a onepage website just
[19:42] for the sake of if I reach out to
[19:45] someone and they want to check out. So
[19:47] like it had a VSSL of me.
[19:49] >> Yeah,
[19:50] >> that's all. And a like a big wall of
[19:53] text below that and it somehow did. And
[19:56] it sent me like a lead every other day.
[20:00] And these leads also like the ones that
[20:02] find you from ChatGPT and Google when
[20:05] they book a call they do show up in as
[20:08] opposed to when you are reaching out
[20:10] people cold call with a cold outreach
[20:12] they might flake on you but not these
[20:14] people.
[20:15] >> Yeah.
[20:16] >> So the quality of the traffic is is very
[20:18] good. It's a really it's a strange
[20:21] phenomenon because for the past what 5,
[20:24] 10, 20 years or so, you go to Google,
[20:27] you bounce around on different websites
[20:30] and you have a a decent feeling about
[20:33] what Google is providing you. You've got
[20:35] reviews there that you may or may not
[20:37] fully trust, right? But there's a lot of
[20:40] trust signals there and you get your top
[20:43] 10 on page one of Google. And now AI has
[20:48] given you three options, usually
[20:50] recommending three, sometimes four I've
[20:52] seen. And those are like the these are
[20:56] the companies. There's nobody else that
[20:57] exists, right?
[20:58] >> Yeah.
[20:59] >> Which is crazy because now you have to
[21:02] figure out how to get as one of those
[21:05] four that are recommended. I did a
[21:08] another episode last week
[21:11] where I was actually analyzing one
[21:15] particular prompt for Phoenix, Arizona,
[21:18] and it related to HVAC repair. And so I
[21:21] put that prompt in. I was using
[21:23] temporary uh chats so that each time it
[21:27] was like a fresh run
[21:28] >> and it was with Gemini and every time it
[21:34] would give me two names that everyone in
[21:36] the entire country knows. These are huge
[21:39] companies. And then one guy that's got
[21:42] 111,
[21:44] not 1100, 111 reviews. And he came up
[21:48] multiple times. I'm going, how is this
[21:50] guy doing this?
[21:51] >> So like he had less reviews than he
[21:53] usually. That's what I
[21:55] >> So So the number one recommended by AI
[21:58] and it was pretty consistent. I ran the
[22:00] prompt six times in the temporary chat
[22:03] and it was like 34,000 reviews was
[22:06] number one.
[22:07] >> Okay.
[22:07] >> The second position was like 14,000
[22:09] reviews. Guy had 111 reviews and it was
[22:13] showing up in like the third or fourth
[22:16] position.
[22:16] >> That's it. It makes no sense, right?
[22:18] This doesn't
[22:20] >> in the SEO world.
[22:23] >> There's something wrong here. We're
[22:24] thinking that there's some black hat
[22:27] situation going on because you shouldn't
[22:29] be able to show up like that.
[22:30] >> Yeah. Yeah, it makes sense.
[22:33] >> And so I actually after I recorded the
[22:36] interview and it would have been fun to
[22:37] do it live, but I called the guy up and
[22:40] I said, "Hey, this sounds weird, right?
[22:42] This is abnormal, but I just did a
[22:45] podcast and I was searching in your area
[22:48] through AI and you kept showing up. What
[22:52] What's going on? What are you doing?"
[22:53] And I was expecting like I got this
[22:55] agency, they specialize in this blah
[22:58] blah blah. They know what's going on.
[23:00] No. His response was I provide really
[23:03] good service.
[23:05] >> That's good, but AI doesn't really
[23:08] that's not why they're ranking you over
[23:11] all these other people that probably
[23:12] provide good service, too.
[23:14] >> Yeah. Yeah.
[23:15] >> And so I dug into his website and there
[23:17] was nothing clear to show why he would
[23:21] be getting represented so well. And and
[23:25] again it his was one of the four
[23:28] companies that were the chosen one.
[23:30] Right. The the other I think there's 750
[23:35] HVAC contractors in the Phoenix area. It
[23:37] didn't exist at that point. Right.
[23:40] So yeah, pretty wild.
[23:43] >> Yeah. But I
[23:47] wouldn't like Gemini also take into
[23:50] account where you are searching from. So
[23:51] like he would be just ranking in that
[23:53] particular neighborhood, right? Not in
[23:55] entire Phoenix, Arizona.
[23:58] >> So I wasn't actually searching by
[23:59] neighborhood. I was just searching
[24:00] Phoenix. So I was just searching by the
[24:03] city name. So it was very
[24:04] >> And you are not And you are not in But
[24:07] let's say someone is in Phoenix and they
[24:10] search for
[24:12] HVAC repair. FE Phoenix then Gemini
[24:16] would probably take into account their
[24:19] location and then present
[24:22] >> it would depend on it may I know it does
[24:25] when you do a search on your phone in
[24:27] Google but these were since they were
[24:32] temporary chats granted yeah I'm in
[24:34] California so I'm not in that location
[24:37] but I it would be curious to see I'd be
[24:40] curious to see what the results would be
[24:42] like if you searched actually chat
[24:46] that's a good question for you. Do the
[24:50] search results vary
[24:53] either for you or for Gemini based on
[24:56] your location.
[24:59] >> Let me check that for you.
[25:08] >> Yes, both can use location. ChatGPT
[25:11] search can use your approximate location
[25:13] from your internet connection and Gemini
[25:15] can use general or precise device
[25:17] location if permitted. A temporary chat
[25:20] doesn't necessarily remove those
[25:22] signals. So your Phoenix runs show
[25:24] repeatability under your conditions, but
[25:26] they don't establish what someone
[25:27] physically in Phoenix would see.
[25:30] >> Okay. So, if I'm not located in Phoenix
[25:34] and I'm running a search for that area,
[25:37] my results may be different than someone
[25:40] in Phoenix. Is that right, Chat?
[25:43] >> Let me check that quickly.
[25:47] >> That's right. They may differ even with
[25:49] the same Phoenix prompt. Location is one
[25:52] possible influence, and answers can also
[25:54] vary between runs. So, we need a
[25:56] comparison to say whether location
[25:58] actually changed yours.
[26:00] Interesting. Okay. In chat, if I had
[26:02] something like if there was a way to set
[26:05] my location with a browser extension
[26:09] uh for Gemini, for instance, would that
[26:12] help to simulate searches in that area?
[26:17] >> Let me check. That
[26:21] >> could help depending on what the
[26:23] extension changes. If it changes the
[26:26] browser's reported coordinates, Gemini
[26:28] would need permission to use that
[26:30] location. Your connection could still
[26:32] indicate California. I treat it as a
[26:34] partial simulation. You could check the
[26:36] location Gemini displays, then compare
[26:38] your results with someone actually in
[26:40] Phoenix using the same prompt and model.
[26:43] >> Yeah. Wow. So, the plot thickens then.
[26:46] >> Mhm.
[26:47] >> That's interesting.
[26:49] So I would wonder now one thing he did
[26:52] say is that his phone calls have
[26:55] increased over the last couple months
[26:58] been getting more calls.
[27:00] >> So
[27:02] >> to me it was an indication that
[27:05] >> it was working for him. He had no idea.
[27:08] >> Yeah,
[27:09] >> it was working for him. Yeah. I mean
[27:12] >> tell him to double Tell him to double
[27:14] down on that.
[27:15] >> Yeah. whatever you're doing, do more of
[27:18] it.
[27:19] >> Um, but yeah, and when I called him, it
[27:22] went to voicemail and then he called me
[27:24] back. It was the owner
[27:26] cell phone number that was on the
[27:27] website. This was like
[27:30] >> it was not an established company
[27:32] whatsoever. That's
[27:33] >> I think that's the opportunity that a
[27:36] lot of people are missing right now
[27:39] that AI search is so new that you could
[27:42] show up like your one page did when
[27:47] other things that you're try that you've
[27:49] been trying to do with Google for such a
[27:51] long time is a struggle.
[27:55] >> Yeah. I think if just like you had
[27:59] interest in marketing and you probably
[28:02] that's why you were able to grow your
[28:04] HVAC business. So if someone a marketing
[28:06] minded person puts their mind to this
[28:09] they can really take advantage of this
[28:11] because agencies most agencies I would
[28:14] say don't really care that much about
[28:17] the results as much as a business owner
[28:21] themselves would do. So yeah that's
[28:25] true.
[28:26] >> Yeah. Yeah, and I think that's actually
[28:28] a good segue into learning a little bit
[28:30] about more more about your business.
[28:33] Tell me more about like how you named
[28:35] the business and I noticed you've got
[28:37] some unique guarantees on your website
[28:40] as well. Talk to us about that.
[28:42] Yeah. So,
[28:45] Honest HVAC Marketing Agency, I named it
[28:48] honest because
[28:51] first of all, it's a good
[28:53] positioning thing that in the sea of
[28:57] HVAC marketing agencies, you need to
[29:00] stand out. I wouldn't say that I'm pro
[29:04] the most honest person in the world. I'm
[29:07] not. But it makes for a good brand
[29:11] positioning. And secondly,
[29:14] yes, we do try to be more honest than
[29:18] the other marketing agencies when it
[29:20] comes to so a lot of marketing agency,
[29:22] they just report on
[29:25] rankings and
[29:28] stuff like that. But we don't do that.
[29:31] Moreover, if we
[29:33] don't provide a service or we are not
[29:35] good at something like we are not good
[29:38] at Google ads. So we tell it honestly to
[29:41] the people. If someone is looking
[29:43] specifically for that we don't want to
[29:45] waste their money. So yeah I'm
[29:50] honest like I find myself to be a little
[29:54] more to be a little more honest than the
[29:57] average guy. So it gels well with my
[30:01] personality, this brand. So that's how I
[30:05] named it this way.
[30:06] >> Excellent. Okay. And and how long ago
[30:10] did you create Honest Hack Marketing
[30:13] Agency?
[30:14] >> Yeah, about a year ago as well, just
[30:16] like you.
[30:16] >> Oh, okay. Yeah. Nice.
[30:18] >> Nice. So, yeah, we're on very similar
[30:21] paths. I
[30:21] >> Yeah,
[30:22] >> that's very cool. That's very cool. And
[30:24] yeah, I noticed that you had a unique
[30:26] guarantee. How did you come up with that
[30:29] guarantee and how receptive are
[30:32] customers to it? That seems like a slam
[30:35] dunk if I'm a business owner.
[30:37] >> To be honest, most people don't even
[30:41] care about the guarantee. Uh it's as if
[30:44] they forget about it when they start
[30:46] doing business with us. Plus like with
[30:50] the help of database reactivation
[30:53] Facebook ads it's pretty easy to deliver
[30:56] on this guarantee but once they see how
[31:01] we work and
[31:04] they get to know us they don't really
[31:08] ask about it just it's just like a foot
[31:10] in the door.
[31:12] >> Yeah. Sure. Okay. Now you mentioned
[31:14] database reactivation that that is a hot
[31:18] topic that's getting thrown around a lot
[31:20] right now. What have you found works
[31:22] best for database? So how are you
[31:25] executing on database reactivation?
[31:30] I would say the
[31:33] usual like tuneups those are the go-to
[31:36] things like it works according to the
[31:39] season of course for different seasons
[31:41] there are different things but the
[31:43] >> low cost offers work the best with
[31:46] database reactivation
[31:48] and what else works well asking for
[31:52] referrals yeah that also works well if
[31:54] you give them some kind of discount or
[31:57] some kind of credit Sure. And is that
[32:00] the reactivation campaigns? Are those
[32:03] email, SMS, outbound? What's typically
[32:07] how's that typically working?
[32:09] >> Yeah, it's just SMS that that has the
[32:11] highest open rate
[32:13] >> and deliverability rate. So, yeah, SMS
[32:17] for sure.
[32:18] >> Excellent. Okay. Okay. Anything else
[32:21] that you want to tell us about Honest
[32:22] Hack Marketing Agency?
[32:25] Probably it would be better if because
[32:30] I'm
[32:33] the listeners wouldn't get too much by
[32:35] listening to me talk about our agency.
[32:38] Probably I could provide them some with
[32:40] some more value by talking about another
[32:43] study that we did
[32:45] >> uh about the Google reviews. So we
[32:48] analyzed
[32:50] 17,000
[32:52] something Google reviews 17,241
[32:56] reviews and the number one the most
[33:01] s not surprising but the best finding
[33:04] out of this was that
[33:06] people
[33:08] the most praised thing about an HVAC
[33:11] business is when a technician
[33:14] explains
[33:16] to the homeowner what's what exactly is
[33:19] wrong with their system? What's the
[33:22] breakdown of the cost that so basically
[33:27] telling them that hey this is what went
[33:29] wrong with your system. So that gives a
[33:31] sense of control to the homeowner. They
[33:34] are not left wondering that hey this
[33:37] person just came checked a few things
[33:39] and now he's quoting us like $1,000.
[33:42] >> Mhm.
[33:43] >> He's explaining to us what went wrong.
[33:45] So we now have more information. We can
[33:48] use this information to doublech check
[33:50] if what he's saying is right or wrong.
[33:53] So that thing gets appreciated a lot as
[33:57] well as when a technician explain also
[34:01] explains the cost breakup what he has
[34:04] just quoted to the homeowner. So it's
[34:07] not just one figure thrown on thrown
[34:11] their way. It's like a specific
[34:14] breakdown that hey this much is for
[34:16] labor, this much is for this part, this
[34:18] much is for this part.
[34:20] >> So
[34:21] >> interesting. Okay. And you said how many
[34:23] total reviews?
[34:24] >> We analyzed 17,241
[34:28] reviews across five
[34:31] different cities, the five most popular
[34:33] cities in the US. And I guess there we
[34:38] analyze about 100 or 120
[34:43] businesses. So the all these Yeah.
[34:46] companies. So 17,000 reviews were from
[34:49] 100 something companies.
[34:51] >> Okay. And the reviews what would be
[34:56] interesting is the average star rating.
[34:59] Was the average star rating for all the
[35:01] businesses that you reviewed pretty
[35:03] high? Was there a minimum don't review a
[35:06] business's reviews if they're under
[35:09] three stars on average? Did you have
[35:12] anything like that or was it just any
[35:14] was it more open?
[35:15] >> Yeah, it was totally open that but the
[35:19] finding was that most of the businesses
[35:22] I would say all of the businesses have
[35:24] more than 4.1 rating overall. Okay,
[35:27] that's what it found. So the edge that
[35:30] businesses have isn't their star rating
[35:32] exactly, but what how many reviews they
[35:34] have in total and what those reviews say
[35:36] because what they say also impacts your
[35:39] Google ranking.
[35:41] >> Yeah. Yeah. And we're finding it to be
[35:44] another trust signal for AI search and
[35:47] generative engine optimization obviously
[35:50] because it's unique content like a
[35:52] human. I would say 95% of the time those
[35:57] reviews are written by humans. I don't
[35:58] think we've gotten to the point where
[36:00] we're having our AI review a company or
[36:04] write a company review for us yet.
[36:06] Although I think that's coming.
[36:08] >> Yeah, that's coming for sure. A lot of
[36:10] these CRM and all they give the ability
[36:14] to oneclick generate a review which one
[36:18] could just copy paste.
[36:19] >> Yeah.
[36:20] >> Yeah. Now, what about responses? Did you
[36:23] measure how many of those reviews were
[36:26] getting responded to?
[36:28] >> Yeah. So, one thing, one of the
[36:30] interesting findings about the responses
[36:34] was that
[36:36] so the fivestar reviews get responded
[36:39] 79.6%
[36:41] of the times, but the one and one star
[36:45] reviews and twostar reviews, they get
[36:48] responded just 64% of the time. So one
[36:50] thing was that
[36:52] >> contractors should reply if not at least
[36:56] as much or even more as they are
[36:59] responding to the fivestar reviews
[37:02] because they can present their side of
[37:05] the story. If someone left a one-star
[37:08] review, they could tell that hey this is
[37:11] what happened and if I'm reading it I
[37:14] could get their side of the story
[37:15] instead of just believing the one star
[37:17] review.
[37:19] Yeah, it's an opportunity and
[37:23] contractors handle one-star reviews
[37:25] wrong most of the time. There's too many
[37:28] of them that are using that response as
[37:31] a way to
[37:34] attack is the wrong word, but get back
[37:36] at the customer that's leaving that that
[37:38] review. And I see it when I'm talking to
[37:42] customers or researching a company that
[37:45] is looking to have their marketing done.
[37:47] And there's just there's so many wrong
[37:49] ways of handling a one-star review. And
[37:53] what people don't realize is that some
[37:55] of your best customers are only looking
[37:58] at the one and two star reviews. That
[38:00] that's the only thing that they're
[38:02] searching before because they want to
[38:04] know if something goes sideways, how are
[38:06] you going to deal with them? Are you
[38:08] gonna attack them personally like you
[38:10] did
[38:10] >> these people call them or not?
[38:14] >> Oh jeez. Yeah, that's pretty rough, but
[38:16] it's out there. There was there was a
[38:18] post last week in one of the groups on
[38:20] Facebook and you know the guy's response
[38:24] to the customer was basically going
[38:27] through like he was in a court of law
[38:29] and detailing every single instance and
[38:33] some of it was a little bit much. He was
[38:35] just going overboard with his response
[38:38] about the situation and he was getting a
[38:40] little he was getting getting a little
[38:42] cute with his response. And if I'm a
[38:45] when I responded to him and I said,
[38:46] "Hey, if I'm the customer here, I don't
[38:48] want to deal with any of this. I don't
[38:50] care if you're right or not at this
[38:52] point. Just your response should be
[38:57] either yes, we screwed up. Here's what
[39:00] we're doing about it." or some of my
[39:02] responses when we would get the
[39:04] occasional one-star review and I took it
[39:06] personally,
[39:07] >> you know, and it was and I would post my
[39:09] cell phone number and say, "Hey, this is
[39:12] going sideways. I understand something's
[39:15] wrong. I can't find your information. I
[39:17] want to talk to you. Call me on my cell
[39:20] phone number." And occasionally I would
[39:22] get people that call and they were
[39:25] shocked that the owner was giving them
[39:27] their phone number to call. And that a
[39:30] lot of these people just want to be
[39:31] heard.
[39:32] >> Yeah.
[39:32] >> And they want to know that you
[39:35] understand their concerns and they it's
[39:37] not necessarily that they just want all
[39:38] their money back. Some people do, but
[39:41] that's not necessarily the case. And
[39:42] >> yeah, most of the times it's just being
[39:44] that they want to get hurt. Even I have
[39:48] seen that
[39:49] >> when if I have had a bad experience, I
[39:53] don't want the money back unless it's a
[39:56] really big amount. I don't want the
[39:57] money back. I just want to make sure
[39:59] that
[40:00] >> this doesn't happen again with anyone
[40:03] else. Yeah. Like on these food delivery
[40:06] apps, they say that, "Hey, do you want a
[40:08] refund?" I say, "No, I don't want a
[40:09] refund. I just wanted to let you know
[40:11] that I had a bad bad experience."
[40:13] >> Yeah. Yeah. So, you've done some pretty
[40:16] some interesting research projects. One
[40:19] on reviews
[40:21] um and then the other on the business
[40:23] websites. Do you have your next research
[40:25] project lined up? Do you know what's
[40:26] coming up in the future?
[40:28] >> No, I've published about four so far and
[40:31] I'm now not going to I'm going to put it
[40:35] on hold for now. Focus on other things
[40:38] so that I can get some business coming
[40:41] in instead of mental masturbating with
[40:44] the thought leadership content.
[40:48] >> Yes. Yes, it can be. It's interesting,
[40:50] but does it really move the needle?
[40:52] Yeah, I understand. That's excellent.
[40:54] chat, did you have any other questions
[40:58] for our guest today?
[41:00] >> Let me check that.
[41:07] >> One last question, Shaurya. For an HVAC
[41:10] owner listening who wants to turn what
[41:12] we've discussed into more booked jobs,
[41:15] what's the first thing you'd have them
[41:16] change this week, and how would you
[41:18] check whether it helped?
[41:25] All right. So, the easiest thing that I
[41:29] would say for an HVAC business owner to
[41:33] get results is like
[41:38] SMS and call all your past customers
[41:40] that you haven't requested a review from
[41:43] and have a system in place that makes it
[41:47] automated from the next time around.
[41:49] Because reviews are the one of the
[41:51] biggest
[41:53] pieces of the puzzle when it comes to
[41:55] Google ranking. And when you get start
[41:58] when you start getting more reviews, you
[41:59] start ranking higher. When you start
[42:01] ranking higher, you start getting more
[42:03] jobs. So yeah, focus on getting more
[42:05] reviews.
[42:08] >> Awesome. Yeah, absolutely. The reviews
[42:11] are critical. the other elements that
[42:13] you talked about being important for the
[42:15] website were that really good
[42:18] information um based on your research
[42:21] reports. So, I appreciate you joining us
[42:24] today. Was there anything else that
[42:25] you'd like to leave us with before we
[42:27] close this out?
[42:32] >> Nothing really. I wouldn't pitch myself
[42:35] when Aaron does the same things. It
[42:38] wouldn't be the nice thing to do. It's
[42:40] quite it's quite all right. No problem.
[42:44] A very respectful guest. I appreciate
[42:46] that. Yes. Okay. This has been a chat
[42:49] with chat. Thanks for joining us today.
[42:52] Great conversation and I hope you can
[42:54] take this and start executing it in your
[42:56] business. We'll see you next time.
[43:01] All right.
[43:06] I've put a short intro and five GEO
[43:09] focused questions on screen centered on
[43:12] how